The goat (Capra hircus) is a perfect animal model for analyzing the transcriptome of milk somatic cells (MSCs), as sufficient numbers of somatic cells in goat milk, i.e., exfoliated epithelial cells, can be obtained using noninvasive methods. RNA integrity and purity are the first and most important parameters qualifying samples for transcriptomic tests and next-generation sequencing, as RNA quality influences experimental results. The aim of this study was to optimize a method for obtaining high-quality RNA from goat MSCs, irrespective of effects like breed, lactation stage, health status (e.g., with or without small ruminant lentivirus [SRLV] infection), or number of somatic cells. Milk samples were obtained from goats of two Polish breeds in various lactation stages and in different parities, and from goats infected and not infected with SRLV. Altogether, 412 MSC samples were examined: 206 using method A with fenozol and 206 using method B with QIAzol. Though the overall purity (measured as absorbance ratios at 260 nm/280 nm and 260 nm/230 nm) of the RNA material was comparable, the average yield of RNA isolated using method A was 11.9 µg, while method B’s average yield was 29.9 µg. Moreover, method B resulted good quality RNA suitable for transcriptome analysis. Results were confirmed by RT-qPCR, using 18S rRNA and RPLP0 as the reference genes. The application of our modified treatment method was successful in obtaining high-integrity samples for transcriptomic or next-generation sequencing analysis. Using a 400 mL milk sample cooled in ice directly after milking, securing the cooling chain process from milking to MSC isolation, and applying method B to isolate RNA, we obtained good RNA quality irrespective of the goats’ breed, lactation stage, parity, milk yield, SRLV infection, and even milk yield and number of somatic cells in milk.