What is the most different language to your native language that you have studied?

[Deinonysus]

My native language is English.

Grammar
English: Mostly governed by word order with some vestigial inflection. No genders/ word classes.
Xhosa: Agglutinative. 15 verb classes (but 6 are just plurals of other classes so really 9).

Vowels
English: Tons of vowels and diphthongs
Xhosa: Same 7 vowels as Italian, I don't think there are any diphthongs

Consonants
English: A decent number of consonants; fairly standard except for the þ/ð sounds. Aspiration is willy-nilly and never changes a word's meaning.
Xhosa: Mostly standard consonants except for three clicks, the "LL" sound from Welsh, a normal guttural sound (as found in German, Hebrew, or Spanish), and an extra-guttural noise that sounds like you're dying. You might think that's the hard part, but it isn't; none of these sounds are too challenging by themselves. But, many consonants (including the clicks) can be unvoiced or voiced, slack-voiced (whatever that means), nasalized, aspirated or unaspirated. These differences are reflected in the spelling and I believe they can often change the meaning of a word, and they can be combined. I haven't studied the language for long and I am nowhere near capable of making all of these distinctions.

Stress/tone
English: Stress accent
Xhosa: Tonal with high, high falling, and low tones.

Phonetic?
English: Nope.
Xhosa: Mostly, although tone isn't reflected in writing.

[nooj]

For me, it was Kyrgyz and Lakota. Maybe even Middle Egyptian.

what can you tell me about these three languages?

[IronMike]

For me, it was Kyrgyz and Lakota. Maybe even Middle Egyptian.

what can you tell me about these three languages?

Kyrgyz was my first Turkic language and I liked it, but oh dear God the familial terms. Enough. I didn't care if I was talking about your younger sister (to a man) or whatever. It killed me. The VSO syntax was no problem (if I recall the syntax correctly). There are resources out there in English and of course even more in Russian.

Lakota was fun. Didn't study it long enough or with a tutor (had a tutor for Kyrgyz) so didn't get very far, but loved hearing some Lakota in movies. Would love to take it up again if I could find an IRL tutor.

Middle Egyptian is just plain fun, and I'm looking forward to the TY folks to come out with their new version. I used Allen's Middle Egyptian and can recommend it wholeheartedly. But I must say that you should be comfortable with linguistics.

[Expugnator]

I tend to spread my choices both towards the different and towards the low-hanging fruits. So I'm picking languages from varied language families:

Mandarin: tones, logographic system, classifiers

Russian: verbs of motions, cases, aspect

Georgian: unique verbal system, ejective consonants, limited ergativity

Estonian: three-level consonant and vowel length distinction, partitives, several verbal tenses

Hebrew: abjad, triliteral roots and I have yet to find more

Indonesian: (it feels rather familiar given the big picture)

[vonPeterhof]

I actually tried to (half-assedly) crunch some numbers on this by using the linguistic typology data from WALS to compare Russian to the three languages I've studied that I thought were most likely to be the most different: Abkhaz, Ainu and Vietnamese (I also really wanted to put in Avar, but there was much less data for it in the database). Basically I downloaded their data for those languages and Russian in Excel, deleted all the lines where there was no value for at least one of the languages and compared how many features ended up being different from the corresponding features in Russian. I was actually surprised that Ainu ended up with the largest number of differences from Russian, with a slight lead of four over Abkhaz. The only differences from Russian it didn't share with either Abhaz or Vietnamese were the lack of lateral consonants, differentiation of nominal and verbal conjunctions, expressing epistemic possibility using neither verbal constructions nor affixes on verbs, and having singular-plural verbal pairs with suppletion.

Although looking at the categories it seems like the similarities between Abkhaz and Russian got artificially inflated a bit through the overemphasis on features related to their consonant inventories and gender systems. For example, in the category "Consonant-Vowel Ratio" both Russian and Abkhaz are classified as "High" (as opposed to "Moderately low" in Ainu and "Low" in Vietnamese), but the "High" for Russian, depending on how you count the phonemes, ranges between 5.5 and 7.8, whereas for Abkhaz the range is between 19.33 and 29 (or 33.5, if you count based on the Bzyp dialect). Whereas in gender Russian and Abkhaz appear similar for having three sex-based genders, but I feel like the fact that in Abkhaz the masculine and feminine genders are reserved for animate nouns is a rather non-trivial distinction (albeit one that makes my job as a learner easier rather than harder).

On the other hand, my own subjective perception of greater familiarity of Ainu might be explained by its similarities to Japanese which no longer strike me as particularly special (small consonant inventory, simple syllable structure, SOV word order, agglutination, uncannily similar phonotactics, etc.). By contrast, Abkhaz is in many ways nothing like any other language I've ever studied, not even the geographically close Georgian and Avar. Here's a good couple of examples:

Санбацеи? "When did I go?"
S- + -anba- + -ts- + -ei
I + when + go + past tense marker

Уанбаигеи? "When did he take you(masc.)?"
W- + -anba- + -i- + -g- + -ei
You(masc.) + when + he + take + past tense marker

So here we have:
-polysynthesis (the meaning of a whole sentence packed into one inseparable word)
-ergativity (note how the intransitive construction starts with the subject while the transitive one starts with the object; both Georgian and Avar have an ergative noun case, but they don't have this for verbal person marking)
-verb roots consisting of just a single consonant (yes, those two are far from the only ones like that; for the record, their dictionary forms are ацара "atsara" and агара "agara", respectively)
-the use of more vowel letters than there are supposedly vowel phonemes in the language (there's an explanation for each and every case, but I forgot most of them )

And that's not even getting into the issues around the language having one of the largest consonant inventories in the world (though to be honest, Avar consonants give me a bit more trouble than Abkhaz ones).

[Iversen]

Irish, followed by Bahasa Indonesia, Latin and English (apart from the myriads of loanwords we have imported from that on). And maybe also Albanian and Dhimotiki and the whole gamut of Slavic languages. But Irish is definitely at place one.

[renaissancemedici]

Turkish, for being altogether different, in sound, grammar, syntax. Even the alphabet was in Arabic not so long ago. Don't let some common words and phrases from being neighbors fool you, it's a very different language.

Hebrew, for absolutely everything. Except for one very unexpected thing: pronunciation. All I have to do is produce greek sounds and lo and behold, wonderful instant Hebrew accent. You don't know how helpful and motivating that is. The Israelis also have a great Greek accent, except for the r sound. But that alphabet... I have never been so confused