Adoption of a streamlined version of the bacterial clustered regular interspersed short palindromic repeat (CRISPR)/Cas9 defense system has accelerated targeted genome engineering. The Streptococcus pyogenes Cas9 protein, directed by a simplified, CRISPR-like single-guide RNA, catalyzes a double-stranded DNA break at a specific genomic site; subsequent repair by end joining can introduce mutagenic insertions or deletions, while repair by homologous recombination using an exogenous DNA template can incorporate new sequences at the target locus. However, the efficiency of Cas9-directed mutagenesis is low in Drosophila melanogaster Here, we describe a strategy that reduces the time and effort required to identify flies with targeted genomic changes. The strategy uses editing of the white gene, evidenced by altered eye color, to predict successful editing of an unrelated gene-of-interest. The red eyes of wild-type flies are readily distinguished from white-eyed (end-joining-mediated loss of White function) or brown-eyed (recombination-mediated conversion to the whitecoffee allele) mutant flies. When single injected G0 flies produce individual G1 broods, flies carrying edits at a gene-of-interest were readily found in broods in which all G1 offspring carried white mutations. Thus, visual assessment of eye color substitutes for wholesale PCR screening of large numbers of G1 offspring. We find that end-joining-mediated mutations often show signatures of microhomology-mediated repair and that recombination-based mutations frequently involve donor plasmid integration at the target locus. Finally, we show that gap repair induced by two guide RNAs more reliably converts the intervening target sequence, whereas the use of Lig4169 mutants to suppress end joining does not improve recombination efficacy.