In 1983 Mielke et al. reported the general contamination of garden soils in the center city area of Baltimore, MD, with Pb, Zn and Cd. The sources included both exterior paint, building demolition wastes, and automotive and stack emissions such that center city garden soils were considerably more contaminated than suburban soils (1000 mg kg^-1 vs. <50 mg kg^-1), but even in rural areas soils could become highly Pb contaminated from paint, lead-arsenate use and from burning trash. Study of garden plant uptake showed the now well known pattern that low growing leafy vegetables accumulated more Pb than other species, and that some root vegetables with expanded hypocotyl (carrot, beet, radish) also accumulated Pb from contaminated soils. But other crops (fruits, grains, tall leafy vegetables such as collards) had little Pb even on highly Pb contaminated soils. The key role of housedust-Pb ingestion by children and the potential of soil Pb to be tracked into homes and ingested by hand-to-mouth play was also becoming recognized at that time. We conducted soil feeding tests with rats which indicated that when soil Pb was higher concentration (10,000 vs. 1000 mg kg^-1), more Pb was absorbed, but that Pb equilibrated in soil was much less bioavailable than soluble Pb salt mixed with soil then added to test diets. Much research has been conducted since that time which clarified that exterior soil comprised much lower risk than dust Pb from interior house paint, or industrial dust Pb deposited in homes. Total removal and replacement of soil near homes had little effect on blood-Pb unless the soil contained well over 1000 ppm Pb. Further, treating Pb contaminated soils with phosphate or high Fe biosolids compost was shown to significantly reduce soil Pb bioavailability to rats, and even to humans in the Joplin field test of soil amendments to reduce soil Pb bioavailability. Codling et al. also demonstrated that carrot accumulated Pb within the xylem which is the center of the edible storage root, allowing significant accumulation of Pb in carrot and similar garden crops. However, testing of soils with an extraction method shown to correlate with the results of soil feeding to humans as part of the Joplin test showed that Pb in long term urban gardens has quite low fractional bioavailability likely due to repeated amendment with manures, composts and phosphate, and tillage and cropping which aid soil Pb inactivation. Most urban gardens can be used safely if precautions are taken to minimize soil transfer to homes, and to restrict growing of low leafy vegetables and herbs, and expanded hypocotyl root vegetables (not potato) when soil Pb exceeds about 500 mg kg^-1. Further, Pb in food has much lower bioavailability to humans than Pb in water.

Garden soil Cd contamination is much less frequent and is usually accompanied by >100 times more Zn such that crops cannot accumulate excessive Cd before Zn phytotoxicity substantially reduces plant yield. And high crop Zn inhibits absorption of crop Cd. Soil As may also comprise concern but is poorly accumulated by crops and has low bioavailability in soils with high Fe oxide levels. The most important first step in deciding about gardening and exposure of children to soils is to test the soil for contaminants such as Pb, Zn, Cd, As, and Cu, general fertility (P, K, Ca, Mg) and pH. With this information, parents can determine if gardening is a risk, and if soil should be remediated to protect children from exposure to their contaminated soil. Because old rural houseside soils, and old orchard soils may have become highly Pb and/or As contaminated, even rural gardeners should test soils before exposing children to the soils and crops. Although adults may be at some risk, the risk to children is so much greater than consideration of garden soil risks to adults is usually irrelevant.