Cladosporium carpophilum, the causal agent of peach scab, overwinters in lesions on 1-year-old twigs, from which conidia infect the developing fruit during spring and early summer. Twig lesions constitute the sole source of initial inoculum; therefore, the mode of dissemination of conidia from such lesions to the fruit is of considerable interest. In a 4-year study, we determined the relative importance of air- versus water-borne conidia and their interaction with different fruit wetness sources (splash, twig runoff, and dew) in a peach orchard with areas that had been treated or not treated with fungicide the previous year. The rareness of scab twig lesions in the previously sprayed trees implied that fruit infection in these trees would occur primarily by airborne conidia from unsprayed trees nearby (located within the same tree row or the adjacent row). In the unsprayed areas, additional infections could occur by short-distance waterborne dissemination of conidia from locally abundant twig lesions via splashing or runoff. Beginning at calyx fall, individual fruit were protected from splash by rain shields, protected from runoff by cotton wicks placed proximal to the peduncle, or left untreated. Rain shields were adjustable, allowing rain or dew to be excluded selectively. Various combinations of the shield and wick treatments were implemented in the previously sprayed and unsprayed areas, and statistical comparison of fruit scab severity between individual treatments by linear contrasts allowed us to untangle the relative contributions of the various sources of inoculum and fruit wetness. Results showed that aerial dissemination of conidia contributed little to fruit scab development, even in the presence of fruit surface wetness caused by splashing, runoff, or dew. In contrast, waterborne conidia contributed considerably and significantly (P < 0.0001) to disease development. This was due primarily to the importance of splash in disseminating conidia from twig lesions to the fruit, given that exclusion of splashing via rain shields decreased disease severity by >90%. Runoff water from the twig to the fruit via the peduncle also contributed to scab development, as evidenced by the fact that exclusion of runoff by cotton wicks reduced disease severity by 31.6 to 44.9%; however, this effect was not always statistically significant. The exclusion of dew did not reduce scab severity (P > 0.4), suggesting that it played a limited role in infection in the presence of other fruit wetness sources.