Source: Dallas Police Department NIBRS quarterly incident files, data through June 30, 2026, extracted on July 2, 2026. Police incident records can be revised after initial release as investigations progress or offenses are reclassified, so recent counts may change slightly in later DPD updates. Violent crime includes aggravated assault (non-family violence), robbery (individual and business), murder, and justifiable homicide; family-violence aggravated assaults and sexual assaults are not included. This definition is narrower than the violent-crime category on DPD's public dashboards, which also counts family-violence aggravated assaults and rape, so totals here run lower than DPD's; year-over-year direction is comparable. Homicide counts include justifiable homicides and run higher than DPD's murder-only tallies (Jan–Jun 2025: 75 vs 64; Jan–Jun 2026: 72 vs 64).

Counting: NIBRS records one row per victim; multi-victim incidents are collapsed to a single incident, so counts are incidents, not victims. An incident with multiple offenses is classified by its most serious offense. Incidents are dated by when they occurred; some DPD public tallies count victims rather than incidents and key on the date an offense was reported, so small differences from DPD's running totals are expected. CPAL's interactive explorer counts on the same basis as this report (incidents, family-violence aggravated assaults excluded); small differences can appear when the explorer's data window updates.

Geography: Incidents are assigned to half-mile grid squares; each square is labeled with its nearest major intersection. Council district assignment uses the center point of each square, so areas straddling a boundary are attributed to one district. Area shares are measured against the full city boundary shown on the maps. On the map, incident dots are shifted by a small random amount (up to roughly 70 feet) so incidents at the same address remain visible as separate dots; dot locations are approximate and should not be read as exact addresses.

Significance: "Significant" increases and decreases are identified with a standard criminological test of whether a year-over-year change is larger than normal variation for an area of that size (Poisson-based; Wheeler 2016). 703 half-mile areas recorded at least one violent incident in Jan–Jun 2025 and were tested for change; areas with no Jan–Jun 2025 incidents are excluded from the test and shown separately on the map. Focus-area thresholds in this half-year report are twice those of the quarterly reports because the comparison window is twice as long.