In the search for improved selective antagonist ligands of the A2B adenosine receptor, which have the potential as antiasthmatic or antidiabetic drugs, we have synthesized and screened a variety of alkylxanthine derivatives substituted at the 1-, 3-, 7-, and 8-positions. Competition for 125I-ABOPX (125I-3-(4-amino-3-iodobenzyl)-8-(phenyl-4-oxyacetate)-1-propylxanthine) binding in membranes of stably transfected HEK-293 cells revealed uniformly higher affinity (<10-fold) of these xanthines for human than for rat A2B adenosine receptors. Binding to rat brain membranes expressing A1 and A2A adenosine receptors revealed greater A2B selectivity over A2A than A1 receptors. Substitution at the 1-position with 2-phenylethyl (or alkyl/olefinic groups) and at N-3 with hydrogen or methyl favored A2B selectivity. Relative to enprofylline 2b, pentoxifylline 35 was equipotent and 1-propylxanthine 3 was >13-fold more potent at human A2B receptors. Most N-7 substituents did not enhance affinity over hydrogen, except for 7-(2-chloroethyl), which enhanced the affinity of theophylline by 6.5-fold to 800 nM. The A2B receptor affinity-enhancing effects of 7-(2-chloroethyl) vs 7-methyl were comparable to the known enhancement produced by an 8-aryl substitution. Among 8-phenyl analogues, a larger alkyl group at the 1-position than at the 3-position favored affinity at the human A2B receptor, as indicated by 1-allyl-3-methyl-8-phenylxanthine, with a K(i) value of 37 nM. Substitution on the 8-phenyl ring indicated that an electron-rich ring was preferred for A2B receptor binding. In conclusion, new leads for the design of xanthines substituted in the 1-, 3-, 7-, and 8-positions as A2B receptor-selective antagonists have been identified.