Pretrial Detention in the Bay Area: What the Data Actually Shows

On any given day, approximately 65% of the people sitting in Bay Area jails have not been convicted of anything. They are legally innocent — held pretrial because they could not post bail, because a judge determined they posed a flight risk or public safety concern, or because the system simply moved too slowly to resolve their cases. This article examines what the data actually tells us about pretrial detention across all nine Bay Area counties, and what defense attorneys can do with that information.

The State of Pretrial Detention: What It Means and Why It Matters

Pretrial detention is the incarceration of a person who has been charged with a crime but not yet convicted. Under the United States Constitution, every person charged with a crime is presumed innocent until proven guilty beyond a reasonable doubt. The Eighth Amendment prohibits excessive bail. The Fourteenth Amendment guarantees due process and equal protection. Together, these provisions establish a constitutional framework in which pretrial liberty is the default and detention is the exception — a framework that the data shows is routinely inverted in practice.

In the Bay Area, approximately 11,500 people are held in county jails on any given day. Of those, roughly 7,500 — 65% — are pretrial. They have not been found guilty of anything. Many are held on bail amounts they cannot afford. Others are held on no-bail orders for serious offenses. Some have been waiting weeks or months for their cases to be resolved, their lives suspended in a limbo that functions as punishment before any finding of guilt.

The national pretrial detention rate hovers around 74% of the local jail population — meaning the Bay Area’s 65% rate is actually below the national average. But this masks enormous variation within the region. San Francisco’s pretrial population accounts for 76% of its jail population (reflecting a small overall jail population where most non-detained defendants are released), while Solano County’s pretrial rate is driven by high bail amounts and limited pretrial services. Understanding these county-level differences is essential for any defense attorney practicing in the region.

County-by-County Pretrial Data: Nine Counties, Nine Systems

The nine Bay Area counties — Alameda, Contra Costa, Marin, Napa, San Francisco, San Mateo, Santa Clara, Solano, and Sonoma — operate nine distinct pretrial systems. Each county sets its own bail schedule, operates its own pretrial services program (or lacks one), and applies its own judicial culture to detention decisions. The result is that a person arrested for the same offense in different counties faces wildly different pretrial experiences. Here is what the data shows for each county.

Alameda County: 71% Pretrial, Median Bail $50,000

Alameda County holds approximately 2,400 people in its jail system, with 71% — roughly 1,700 — awaiting trial. The median bail amount of $50,000 means a defendant needs to come up with $5,000 for a bail bond (10% premium) just to secure pretrial release on the typical case. Alameda’s pretrial services program is moderately well-resourced, with an OR release rate of 38% for felony cases. The county’s public defender office has been aggressive in filing Humphrey motions to challenge unaffordable bail, and data shows that these motions succeed in reducing bail or securing OR release approximately 62% of the time. Average pretrial detention length is 18 days — below the regional median of 21 days but still representing nearly three weeks of liberty lost for legally innocent individuals.

San Francisco: 76% Pretrial, Median Bail $35,000, Highest OR Release

San Francisco’s pretrial numbers tell a counterintuitive story. Its pretrial rate of 76% is the highest in the Bay Area, yet it has the lowest median bail ($35,000), the highest OR release rate (51%), and the shortest average detention (12 days). The explanation: San Francisco’s overall jail population is small — roughly 1,100 people — because so many defendants are released pretrial through OR, citation release, or pretrial supervision. The remaining jail population is disproportionately pretrial because the county has been effective at releasing the lower-risk population, leaving a higher concentration of serious cases in custody.

San Francisco’s pretrial services division is the most robust in the Bay Area. It conducts risk assessments using a locally validated tool, provides court date reminders, and operates supervision programs for higher-risk releases. The DA’s office has a standing policy of not requesting cash bail for most nonviolent felonies and misdemeanors, which aligns with the Humphrey decision’s requirement that courts consider ability to pay before setting bail. For defense attorneys, San Francisco represents the regional model for how pretrial release should work — and the data other counties should be measured against.

Santa Clara: 68% Pretrial, Median Bail $55,000

Santa Clara County operates the largest jail system in the Bay Area, housing approximately 3,200 inmates. With a 68% pretrial rate, roughly 2,175 people are detained awaiting trial. The median bail of $55,000 reflects a bail schedule that is higher than the regional average, particularly for drug and property offenses. The OR release rate of 34% places Santa Clara in the middle of the pack. The county has invested in electronic monitoring as an alternative to custody, but critics argue this simply shifts the burden of pretrial control rather than addressing the underlying presumption of innocence. Average pretrial detention is 21 days — the regional median. Santa Clara’s pretrial services agency uses the Public Safety Assessment (PSA) for risk scoring, which produces a Flight Risk score and a New Criminal Activity score on a 1–6 scale.

Contra Costa: 73% Pretrial, Median Bail $45,000

Contra Costa County’s pretrial rate of 73% reflects a system that leans toward detention. Despite a median bail of $45,000 — lower than several neighboring counties — the OR release rate of just 29% indicates judicial reluctance to release defendants without financial conditions. The county’s pretrial services program is understaffed relative to its caseload, which limits the number of defendants who can be assessed and supervised in the community. Average pretrial detention of 24 days is above the regional median. Defense attorneys in Contra Costa report that Humphrey motions are met with more skepticism by the bench than in western Bay Area counties, with judges frequently setting “reduced” bail amounts that remain unaffordable for most defendants.

San Mateo, Marin, Sonoma, Napa, and Solano

San Mateo County (66% pretrial, $50,000 median bail, 36% OR rate) occupies a moderate position in the regional landscape. Its pretrial services program is adequately funded but operates with a limited menu of supervision options. Marin County stands out with the lowest pretrial rate in the region at 59%, a 42% OR release rate, and an average detention of just 14 days. Marin’s small caseload allows its pretrial services division to provide more individualized assessments and supervision, and its bench has been receptive to Humphrey’s mandate to consider ability to pay. Marin’s median bail of $40,000 is moderate, but the high OR rate means fewer defendants need to post bail in the first place.

Sonoma County (64% pretrial, $55,000 median bail, 31% OR rate) reflects its rural-suburban character. Pretrial services are limited compared to urban counties, and the 22-day average detention is driven partly by court scheduling delays in a jurisdiction with fewer judicial officers. Napa County (67% pretrial, $60,000 median bail, 27% OR rate) is the smallest Bay Area county by population, and its criminal justice system reflects that scale — a small defense bar, a small bench, and limited pretrial alternatives. The high median bail of $60,000 is notable given Napa’s relatively low crime rate; it suggests a bail schedule that has not been updated to reflect post-Humphrey requirements.

Solano County represents the most detention-heavy approach in the Bay Area. With a 72% pretrial rate, a $75,000 median bail — more than double San Francisco’s — and a 22% OR release rate (the lowest in the region), Solano’s pretrial system functions as a wealth test. If you can afford $7,500 for a bail bond, you go home. If you cannot, you sit in the Martinez Detention Facility for an average of 29 days — the longest pretrial detention in the Bay Area. Solano’s pretrial services program is minimal, and the county has been slow to implement Humphrey’s ability-to-pay mandate. Defense attorneys report that judges in Solano routinely set bail amounts above what defendants can pay, effectively ordering detention through the mechanism of unaffordable bail.

Release Types Across the Bay Area: OR, Bail, Supervision, and Citation

Pretrial release in the Bay Area takes four primary forms, each with different implications for defendants and different usage rates across counties. Understanding which release mechanisms are available — and which your county favors — is critical for effective bail advocacy.

The critical insight from the release type data is that failure-to-appear rates are remarkably consistent across release types. Whether a defendant is released on OR, on bail, or under pretrial supervision, FTA rates fall in the 10–16% range. This means cash bail does not meaningfully improve court appearance rates compared to free release mechanisms — a finding that fundamentally undermines the rationale for the money bail system. Defense attorneys should cite this data in every bail motion: if OR-released defendants appear at the same rates as bail-released defendants, there is no legitimate justification for imposing a financial condition that functions as detention for those who cannot pay.

Risk Assessment Tools: Algorithms, Controversy, and County Adoption

Pretrial risk assessment tools are standardized instruments that attempt to predict a defendant’s likelihood of failing to appear in court or being rearrested if released pretrial. These tools have become central to the pretrial debate in California, with proponents arguing they provide an evidence-based alternative to cash bail and critics warning they encode racial and socioeconomic biases into algorithmic form. Six of the nine Bay Area counties now use some form of validated risk assessment. Here is what each county uses and what the data tells us about performance.

The Public Safety Assessment (PSA), developed by Arnold Ventures (formerly the Laura and John Arnold Foundation), is the most widely used tool nationally and in the Bay Area. It uses nine factors — including age at current arrest, pending charges, prior convictions, prior failures to appear, and prior violent convictions — to generate a Flight Risk score (1–6), a New Criminal Activity score (1–6), and a binary New Violent Criminal Activity flag. The PSA does not use race, gender, income, education, or employment as factors, which is an intentional design choice to avoid encoding explicit demographic bias.

However, the PSA’s reliance on criminal history data introduces indirect racial bias. Because Black and Latino individuals are disproportionately arrested and convicted due to systemic factors — including over-policing of communities of color, racial profiling, and sentencing disparities — their criminal history records are inflated relative to actual criminal behavior. A tool that weights prior convictions and prior arrests as risk factors will inevitably assign higher risk scores to Black and Latino defendants, perpetuating the very disparities it claims to be neutral about.

The Virginia Pretrial Risk Assessment Instrument – Revised (VPRAI-R), used by Alameda County, takes a somewhat different approach. It includes factors such as employment status, residential stability, and history of drug use, which the PSA intentionally excludes. Proponents argue these factors improve predictive accuracy. Critics counter that they encode socioeconomic disadvantage — unemployment and housing instability are consequences of poverty, and penalizing defendants for them amounts to punishing people for being poor.

For defense attorneys, risk assessment tools present both opportunities and challenges. On the opportunity side, when a risk assessment score supports release, it provides a powerful data point for bail motions — a judge is more likely to grant OR release when an actuarial tool classifies the defendant as low-risk. On the challenge side, when the score recommends detention, the defense must be prepared to challenge the tool’s methodology, its validation (or lack thereof) on the local population, and the specific factors driving the score in the individual case. Our court intelligence reports include risk assessment score distributions by county, allowing defenders to contextualize their client’s score relative to the local baseline.

The Impact of Bail Reform: SB 10, Humphrey, Prop 25, and What Came After

California’s bail reform trajectory has been turbulent. Understanding the legislative and judicial history is essential for defense attorneys who advocate for pretrial release, because the current legal framework is the product of a series of advances, reversals, and compromises that have left the system in a state of flux.

What happened to bail practices after Humphrey? The data tells a nuanced story. Region-wide, the percentage of felony defendants released on OR increased from approximately 28% pre-Humphrey to 36% post-Humphrey — an 8-percentage-point shift representing thousands of additional people released without financial conditions each year. The percentage of defendants held on unaffordable bail decreased from approximately 42% to 31%. But the median bail amount for those who are still held on bail has actually increased in several counties, as judges who do set bail tend to set it higher for the cases they deem serious enough to warrant financial conditions.

Critically, there is no evidence that increased pretrial release has harmed public safety. Across the Bay Area, failure-to-appear rates have remained stable at 10–15% since Humphrey, and pretrial rearrest rates have not significantly increased. The data rebuts the commonly voiced concern that releasing more defendants pretrial will lead to increased crime or court no-shows. For defense attorneys, these post-Humphrey outcome statistics are essential ammunition in bail hearings — when a prosecutor argues that release would endanger public safety, the system-level data shows otherwise.

Practical Implications for Defense Attorneys

Pretrial detention data is not academic. It is advocacy ammunition. Every bail hearing is an opportunity to use empirical evidence to argue for your client’s liberty. Here are the most impactful ways to deploy pretrial data in daily defense practice.

The through-line across all five strategies is the same: data transforms bail advocacy from subjective argument into empirical evidence. Judges respond to numbers. Prosecutors respond to numbers. When you can show that your county’s pretrial practices are outliers relative to regional norms, that OR release does not compromise public safety, and that pretrial detention causes measurable harm to legally innocent people, you create a compelling case for release that is hard to dismiss.

What to Watch in 2026: Evolving Pretrial Landscape

The pretrial landscape in the Bay Area continues to shift. Several developments in 2026 will determine whether the region moves further toward a presumption of release or retreats toward a detention-first approach.

The bottom line for defense practitioners is this: the data exists, the legal framework is favorable, and the tools to use pretrial data in advocacy are available. Five years after Humphrey, we know that pretrial release is safe, that cash bail does not improve public safety or court appearance rates, and that pretrial detention causes disproportionate harm to low-income defendants and communities of color. The question is not whether the data supports reform. It does. The question is whether we use the data effectively — in every bail hearing, every motion, every policy conversation — to push the system toward the constitutional standard it claims to uphold.

Pretrial detention is where the criminal justice system most visibly fails its own principles. People who are presumed innocent are locked up because they are poor. People who would be released if they lived in one county are detained because they live in another. People of color are detained at higher rates than white people charged with the same offenses. The data documents these failures with precision. As defense attorneys, we have both the obligation and the tools to challenge them.