We describe, for the first time the use of hydrogel-forming microneedle (MN) arrays for minimally-invasive extraction and quantification of drug substances and glucose from skin in vitro and in vivo. MN prepared from aqueous blends of hydrolysed poly(methyl-vinylether-co-maleic anhydride) (11.1% w/w) and poly(ethyleneglycol) 10,000 daltons (5.6% w/w) and crosslinked by esterification swelled upon skin insertion by uptake of fluid. Post-removal, theophylline and caffeine were extracted from MN and determined using HPLC, with glucose quantified using a proprietary kit. In vitro studies using excised neonatal porcine skin bathed on the underside by physiologically-relevant analyte concentrations showed rapid (5 min) analyte uptake. For example, mean concentrations of 0.16 μg/mL and 0.85 μg/mL, respectively, were detected for the lowest (5 μg/mL) and highest (35 μg/mL) Franz cell concentrations of theophylline after 5 min insertion. A mean concentration of 0.10 μg/mL was obtained by extraction of MN inserted for 5 min into skin bathed with 5 μg/mL caffeine, while the mean concentration obtained by extraction of MN inserted into skin bathed with 15 μg/mL caffeine was 0.33 μg/mL. The mean detected glucose concentration after 5 min insertion into skin bathed with 4 mmol/L was 19.46 nmol/L. The highest theophylline concentration detected following extraction from a hydrogel-forming MN inserted for 1 h into the skin of a rat dosed orally with 10 mg/kg was of 0.363 μg/mL, whilst a maximum concentration of 0.063 μg/mL was detected following extraction from a MN inserted for 1 h into the skin of a rat dosed with 5 mg/kg theophylline. In human volunteers, the highest mean concentration of caffeine detected using MN was 91.31 μg/mL over the period from 1 to 2 h post-consumption of 100 mg Proplus^® tablets. The highest mean blood glucose level was 7.89 nmol/L detected 1 h following ingestion of 75 g of glucose, while the highest mean glucose concentration extracted from MN was 4.29 nmol/L, detected after 3 hours skin insertion in human volunteers. Whilst not directly correlated, concentrations extracted from MN were clearly indicative of trends in blood in both rats and human volunteers. This work strongly illustrates the potential of hydrogel-forming MN in minimally-invasive patient monitoring and diagnosis. Further studies are now ongoing to reduce clinical insertion times and develop mathematical algorithms enabling determination of blood levels directly from MN measurements.