The effect of prior malignancy on the risk of developing, and prognosis of, acute lymphoblastic leukemia (ALL) is unknown. This observational study utilized the California Cancer Registry to estimate the risk of developing ALL after a prior malignancy using standardized incidence ratios (SIRs, 95% confidence intervals). ALL occurring after a malignancy with an SIR>1 (increased-risk (IR) malignancies) was considered secondary ALL (s-ALL). Adjusted hazard ratios (aHRs, 95% confidence intervals) compared the effect of s-ALL with de novo ALL on overall survival. A total of 14 481 patients with ALL were identified (1988-2012) and 382 (3%) had a known prior malignancy. Any prior malignancy predisposed patients to developing ALL: SIR 1.62 (1.45-1.79). Hematologic malignancies (SIR 5.57, 4.38-6.98) and IR-solid tumors (SIR 2.11, 1.73-2.54) increased the risk of developing ALL. s-ALL increased the risk of death compared with de novo ALL (aHR 1.38 (1.16-1.63)) and this effect was more pronounced among younger patients (age<40 years: aHR 4.80 (3.15-7.30); age⩾40 years: aHR 1.40 (1.16-1.69)) (interaction P<0.001). This population-based study demonstrates that s-ALL is a distinct entity that occurs after specific malignancies and carries a poor prognosis compared with de novo ALL, particularly among patients <40 years of age.