Sixteen 5-aryl-substituted isothiazol-3(2H)-one-1,(1)-(di)oxide analogs have been prepared from the corresponding 5-chloroisothiazol-3(2H)-one-1-oxide or −1,1-dioxide by a Suzuki-Miyaura cross-coupling reaction and screened for their inhibition potency against four human carbonic anhydrase isoenzymes: the transmembrane tumor-associated hCA IX and XII and the cytosolic off-target hCA I and II. Most of the synthesized derivatives inhibited hCA IX and XII isoforms in nanomolar range, whereas remained inactive or modestly active against both hCA I and II isoenzymes. In the N-tert-butylisothiazolone series, the 5-phenyl-substituted analog (1a) excelled in the inhibition of tumor-associated hCA IX and XII (Ki = 4.5 and Ki = 4.3 nM, respectively) with excellent selectivity against off target hCA I and II isoenzymes (S > 2222 and S > 2325, respectively). Since the highest inhibition activities were observed with N-tert-butyl derivatives, lacking a zinc-binding group, we suppose to have a new binding mode situated out of the active site. Additionally, three free-NH containing analogs (3a, 4a, 3i) have also been prepared in order to study the impact of free-NH containing N-acyl-sulfinamide- (-SO-NH-CO-) or N-acyl-sulfonamide-type (-SO2-NH-CO-) derivatives on the inhibitory potency and selectivity. Screening experiments evidenced 5-phenylisothiazol-3(2H)-one-1,1-dioxide (4a), the closest saccharin analog, to be the most active derivative with inhibition constants of Ki = 40.3 nM and Ki = 9.6 nM against hCA IX and hCA XII, respectively. The promising biological results support the high potential of 5-arylisothiazolinone-1,(1)-(di)oxides to be exploited for the design of potent and cancer-selective carbonic anhydrase inhibitors.