This paper studies nonparametric identification of nonseparable duration models with unobserved heterogeneity. The models considered here are nonseparable in two ways. First, genuine duration dependence is allowed to depend on observed covariates. Second, observed and unobserved characteristics may interact in an arbitrary way. Identification is shown for a comprehensive account of settings. In particular, identification is shown in single-spell models with and without time-varying covariates, in multiple-spells models with shared frailty and lagged duration dependence, in single-spell and multiple-spell competing risks models, and in treatment effects models where treatment is assigned during the individual spell in the state of interest.