Covertly attending to a stimulus location increases spatial acuity. Is such increased spatial acuity coupled with a decreased acuity at unattended locations? We measured the effects of exogenous (transient, involuntary) and endogenous (sustained, voluntary) attention on observers’ acuity thresholds for a Landolt gap resolution task at both attended and unattended locations. Both types of attention increased acuity at the attended and decreased it at unattended locations relative to a neutral baseline condition. These trade-off findings support the idea that limited processing resources affect early vision, even when the display is impoverished and there is no location uncertainty. There was no benefit without a cost.