Tamoxifen (Tm)-inducible Cre recombinases are widely used to perform gene inactivation and lineage tracing studies in mice. Although the efficiency of inducible Cre-loxP recombination can be easily evaluated with reporter strains, the precise length of time that Tm induces nuclear translocation of CreER^Tm and subsequent recombination of a target allele is not well defined, and difficult to assess. To better understand the timeline of Tm activity in vivo, we developed a bioassay in which pancreatic islets with a Tm-inducible reporter (from Pdx1^PB-CreER^Tm;R26R^lacZ mice) were transplanted beneath the renal capsule of adult mice previously treated with three doses of 1 mg Tm, 8 mg Tm, or corn oil vehicle. Surprisingly, recombination in islet grafts, as assessed by expression of the β-galactosidase (β-gal) reporter, was observed days or weeks after Tm treatment, in a dose-dependent manner. Substantial recombination occurred in islet grafts long after administration of 3×8 mg Tm: in grafts transplanted 48 hours after the last Tm injection, 77.9±0.4% of β-cells were β-gal+; in β-cells placed after 1 week, 46.2±5.0% were β-gal+; after 2 weeks, 26.3±7.0% were β-gal+; and after 4 weeks, 1.9±0.9% were β-gal+. Islet grafts from mice given 3×1 mg Tm showed lower, but notable, recombination 48 hours (4.9±1.7%) and 1 week (4.5±1.9%) after Tm administration. These results show that Tm doses commonly used to induce Cre-loxP recombination may continue to label significant numbers of cells for weeks after Tm treatment, possibly confounding the interpretation of time-sensitive studies using Tm-dependent models. Therefore, investigators developing experimental approaches using Tm-inducible systems should consider both maximal recombination efficiency and the length of time that Tm-induced Cre-loxP recombination occurs.