Aptamers are oligonucleotide-based affinity reagents that are increasingly being used in various applications. Systematic evolution of ligands by exponential enrichment (SELEX) has been widely used to isolate aptamers for small-molecule targets, but it remains challenging to generate aptamers with high affinity and specificity for targets with few functional groups. To address this challenge, we have systematically evaluated strategies for optimizing the isolation of aptamers for (+)-methamphetamine, a target for which previously reported aptamers have weak or no binding affinity. We perform four trials of library-immobilized SELEX against (+)-methamphetamine and demonstrate that N30 libraries do not yield high-quality aptamers. However, by using a more complex N40 library design, stringent counter-SELEX, and fine-tuned selection conditions, we identify aptamers with high affinity for (+)-methamphetamine and better selectivity relative to existing antibodies. Bioinformatic analysis from our selections reveals that high-quality aptamers contain long conserved motifs and are more informationally dense. Finally, we demonstrate that our best aptamer can rapidly detect (+)-methamphetamine at toxicologically relevant concentrations in saliva in a colorimetric dye-displacement assay. The insights provided here demonstrate the challenges in generating high-quality aptamers for low complexity small-molecule targets and will help guide the design of more efficient future selection efforts.