Accidentally introduced parasitoids can provide biological control of the invasive pest they followed but could also pose a threat to native species if they are not specialists. Attack of native species could reduce the effec- tiveness of adventive natural enemies if it is accompanied by reduced fitness, and it may increase acceptance of non-target species. We tested how intermittent host shift and prolonged exposure to a non-target, native species Podisus maculiventris (Say) (Hemiptera: Pentatomidae) may affect the fitness and host acceptance of the egg parasitoid, Trissolcus japonicus (Ashmead) (Hymenoptera: Scelionidae) that was co-introduced with the invasive Halyomorpha halys (Stål) (Hemiptera: Pentatomidae). In the first experiment, replicate parasitoid populations were reared either with or without a history of host switching over a 10-generation period between the native and the invasive stink bug that was followed by three generations of continuous reproduction on either host. In the second experiment, replicated populations were exposed for three continuous generations to either the native or invasive host. Occasional or prolonged reproduction on the native host did not increase acceptance and developmental success on the non-target species. Emergence rate of T. japonicus was consistently lower when it reproduced on the native stink bug. There was also a trade-off for the parasitoid, in terms of reduced emergence rate, when switching back to H. halys from the native host. These results suggest that host shifts to native species could reduce the effectiveness of T. japonicus at attacking H. halys, but the fitness costs may prevent significant non-target effects in nature.